Prevalence and rates of recognition of anxiety disorders in internal medicine outpatient departments of 23 general hospitals in Shenyang, China

Available online at www.sciencedirect.com

General Hospital Psychiatry 32 (2010) 192 – 200

Prevalence and rates of recognition of anxiety disorders in internal medicine outpatient departments of 23 general hospitals in Shenyang, China Xiaoxia Qin, M.Med.a,⁎, Michael R. Phillips, M.D., M.P.H.b,c , Wei Wang, M.Med.a , Yueling Li, M.D.a , Qiu Jin, M.Med.a , Li Ai, M.Med.a , Shengnan Wei, M.Med.a , Guanghui Dong, Ph.D.d , Li Liu, M.D.a a

Department of Psychiatry, First Affiliated Hospital, China Medical University, Shenyang, Liaoning 110001, China b Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai 200030, China c Departments of Psychiatry and Global Health, Emory University School of Medicine, Atlanta, GA 30322, USA d Statistics Center, School of Public Health, China Medical University, Shenyang, Liaoning 110001, China Received 13 July 2009; accepted 1 December 2009

Abstract Objective: Assess the prevalence, risk factors and treating clinicians' rates of recognition of anxiety disorders in internal medicine departments of different types of general hospitals in Shenyang, China. Method: A two-stage screening process using an expanded Chinese version of the 12-item General Health Questionnaire (GHQ) and the Structured Clinical Interview for DSM-IV-TR Axis I Disorders (SCID) identified 457 persons 15 years of age or older with current anxiety disorders from among 5312 consecutive attendees at the outpatient internal medicine departments of 23 randomly selected general hospitals. Clinical charts were reviewed to determine whether or not the treating internist had made a diagnosis of anxiety or prescribed anxiolytic medications. Results: The 1-month prevalence of any type of anxiety disorder was 9.8% (95% CI=9.0–10.8%). The prevalences of the three most common disorders: anxiety disorder not otherwise specified, generalized anxiety disorder and anxiety disorder due to a general medical condition, were 6.3% (5.6–7.1%), 2.4% (2.0–2.9%), and 0.6% (0.4–0.8%), respectively. Multivariate logistic regression analysis identified the following independent predictors of having a current anxiety disorder: every being married (OR=3.5, 95% CI=2.3–5.4), prior treatment for psychological problems (3.3, 1.8–6.0), having religious beliefs (1.9, 1.3–2.9), low family income (1.5, 1.2–1.9) and never having attended college (1.3, 1.02–1.8). Among the 402 patients with anxiety disorders for whom the clinical chart was reviewed only 16 (4.0%, CI=2.3–6.3%) were diagnosed with an anxiety condition or treated with anxiolytic medications. Conclusion: The prevalence of anxiety in internal medicine outpatients in urban China is lower than that reported in most western countries and the profile of risk factors is somewhat different. The very low rate of recognition of these disorders by internists is related both to the low rates of care-seeking for psychological problems in the general population and to the high-volume collective model of care delivery in the outpatient departments of Chinese general hospitals. Steps to increase the recognition and treatment of anxiety disorders in Chinese general hospitals must focus both on changing attitudes of patients and clinicians and, more importantly, on altering the structure of care delivery. © 2010 Elsevier Inc. All rights reserved. Keywords: General hospital; Anxiety; Prevalence; Rate of recognition

1. Introduction

⁎ Corresponding author. Tel.: +86 24 8328 2184. E-mail address: [email protected] (X. Qin). 0163-8343/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.genhosppsych.2009.12.001

Since the publication of the Global Burden of Disease studies starting in 1996 [1–3], it has been recognized that the overall effect on community health of common mental disorders such as depression and anxiety is much greater

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

than that of more severe but less prevalent mental disorders such as schizophrenia [4]. The majority of individuals suffering from these conditions never seek treatment for their psychological problems so there have been continuing efforts to increase the recognition and treatment of common mental disorders in primary care settings [5,6]. Most available research on depression and anxiety in primary care settings comes from high-income countries, but over the last decade increased attention has focused on the prevalence, recognition and management of common mental disorders in the primary care settings of low- and middleincome countries [7–11]. One major limitation of prior reports of primary care clinicians' recognition rates for mental disorders [12,13] is that the typical method of assessing recognition—requiring clinicians to record the possibility of the disorder(s) as they treat their patients—probably leads to a substantial inflation of the actual recognition rate. Requiring clinicians to record the presence or absence of a specific mental disorder is a significant departure from routine care that assesses whether or not primary care clinicians are capable of making the corresponding diagnosis, it does not assess the frequency with which they do so in their routine care. Another limitation of studies on the recognition and management of anxiety disorders in general health care settings is that they are typically restricted to specific disorders such as generalized anxiety disorder (GAD) [14,15] or posttraumatic stress disorder [16,17]. Few studies consider the common form of anxiety seen in general health care—the subsyndromal form of anxiety which is assigned the DSM-IV diagnosis of anxiety disorder not otherwise specified (NOS) [18]. The prevalence of anxiety disorders in mainland China is somewhat lower than that reported in developed countries but it is, nevertheless, an important public health problem. A recent large epidemiological study in four provinces of China [19] reported a current combined prevalence of DSM-IV anxiety disorders in adults of 5.6% (95% CI=5.0–6.3). Importantly, among all persons with anxiety disorders 59% had anxiety disorder NOS, 23% were moderately to severely disabled by their condition, and only 6% had ever sought any type of professional help. These low rates of care-seeking confirm the importance of primary health care settings as key venues in the international effort to improve the identification and treatment of common mental disorders [10,11]. In urban China, which now accounts for approximately half of the national population, the majority of primary care services are provided in outpatient clinics of general hospitals. Thus, a crucial first step in the management of common mental disorders is to obtain accurate information about the prevalence of these conditions among patients seen in general hospitals and about the proportion of those with mental disorders that are currently identified and treated. To help clarify these issues, this study identified a large sample of outpatients at internal medicine departments from a representative group of 23 general hospitals in Shenyang,

193

China (a city of 6.9 million persons in northeastern China), determined the presence of various mental disorders in these subjects and then reviewed the clinical charts of the patients to determine whether or not the treating clinician had diagnosed or provided treatment for any mental disorder. A previous report [20] discussed the results of the study for depressive disorders. In this article, we present the results for anxiety disorders.

2. Methods 2.1. Sampling The 96 general hospitals in the Shenyang municipality were subdivided into 47 primary hospitals (less than 100 inpatient beds), 31 secondary hospitals (100–500 beds) and 18 tertiary hospitals (over 500 beds) and then 20% of each type of hospital was randomly selected using a random number table. This resulted in 23 target hospitals: 10 primary, 7 secondary, and 6 tertiary. The total target sample of 5750 internal medicine outpatients was subdivided as follows: 990 from primary general hospitals, 2056 from secondary hospitals and 2704 from tertiary hospitals. 2.2. Procedures The research team sequentially visited each of the 23 target hospitals from November 2004 through January 2006. Identified subjects were sequential patients 15 years of age or older being treated in the internal medicine outpatient department of each hospital who were physically able to complete the interview. Prior to seeing the internal medicine clinician, patients were asked to provide written informed consent to participate in the project. If the patient agreed, demographic information was obtained and an expanded Chinese version of the 12-item General Health Questionnaire (GHQ-12) [21] was administered by a trained interviewer. Subjects were classified as high-risk, moderaterisk or low-risk for mental disorder based on the GHQ results and all high-risk subjects, a random selection of 40% of moderate-risk subjects and a random selection of 10% of low-risk subjects were administered the full Structured Clinical Interview for DSM-IV-TR Axis I Disorders (SCID) [22]. The patient then saw the internist and, if the SCID exam had identified any psychiatric disorder, the patient's chart (which is always in the patient's possession) was examined after the clinical interview to determine whether or not the internist made a diagnosis of anxiety and/or prescribed anxiolytic medications. Treating internists remained blind to the purpose of the study and to the SCID diagnosis so they did not change their assessment and management of patients in any way during the study; this ensured that the recognition rates identified were not inflated due to study-related changes in routine care. Patients who were diagnosed with any current mental illness by the SCID examination but not identified by the

194

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

treating internists were informed of the diagnosis after seeing the internists and advised to seek treatment at a specialty mental health service. 2.3. Measures The screening instrument included the 12-item GHQ, which has previously been validated in China by Yang et al. [23], and eight additional items aimed at increasing the sensitivity of the screening process: (1) subjective report of very poor physical health in last month, (2) subjective report of very poor psychological health in last month, (3) obsessive thoughts or compulsive behaviors over last month, (4) restriction of behavior because of phobias over last month, (5) high levels of nervousness or anxiety over last 6 months, (6) social problems due to drinking over last year, (7) medical treatment for psychological problems at any time in past and (8) prior psychiatric hospitalization. Each of the 12 GHQ items has a four-answer response set and one point is assessed if the respondent reported either one of the two more pathological answers for the item; this results in a 0–12 overall score. The eight risk additional items were coded as dichotomous (present/absent) variables. High-risk patients were those with any of the eight risk factors present or a GHQ score of 5 or above; moderate-risk patients had no risk factors and a GHQ score of 2–4; low-risk patients had no risk factors and a GHQ score of 0 or 1. The screening interviews were conducted by four psychiatric nurses and three masters-level research fellows who received one week training in the screening process. Over the course of the study, 415 randomly selected subjects were re-administered the GHQ by a second interviewer who was blind to the first result; the inter-rater reliability for the three risk levels was excellent (intraclass correlation coefficient [ICC]=0.996). Patients' diagnoses were determined by psychiatrists who administered a revised Chinese version of the SCID [24,25] and who were blind to the results of the GHQ screening. In this study, “any anxiety disorder” included panic disorder, social phobia, specific phobias, obsessive– compulsive disorder, posttraumatic stress disorder (PTSD), GAD, anxiety disorder due to a general medical condition, and anxiety disorder NOS. The clinical exam typically took about 50 min. The six psychiatrists who participated in the study attended a four-week training course in the use of the SCID; their inter-rater reliability at the end of training using 16 taped interviews of different types of patients was excellent (ICC=0.95). “Recognized” cases were those in which the SCID exam resulted in a diagnosis of any anxiety disorder and the treating internists recorded “anxiety” or some related term as a diagnosis in the patient's chart and/or prescribed any type of anxiolytic medication. Among the 457 cases with any SCID anxiety disorder diagnosis, 55 (12%) left the clinic before it was possible to review the chart so the recognition rate estimates were limited to 402 cases. Patients with

anxiety disorders for whom the internist's record was not available were younger than those for whom it was available [mean (S.D.) age=45 (16) vs. 51 (14), t=2.53, P=.012], but there were no significant differences between the two groups in gender, educational level or family income. 2.4. Statistical methods Both the primary and secondary diagnoses were considered in this analysis. Adjusted rates of disorders were weighted by the proportion of SCID examinations completed in the high-, medium- and low-risk subjects (i.e., the inverse of the sampling fraction inflated by the number of scheduled but not completed SCID interviews) and the 95% CIs were computed using the Gaussian approximation to the log likelihood of the log (rate) [26]. Comparison of the characteristics of subjects with an anxiety diagnosis to those without a diagnosis of anxiety also required adjustment for the different numbers of subjects in the three risk groups to ensure that the comparison related to characteristics of anxiety in all internal medicine outpatients (and not just in those who received a SCID exam). In the comparison of those with and without an anxiety disorder some variables were dichotomized for ease of interpretation: family income was categorized as above and below $1250 per year; educational status was dichotomized as with or without some college education; occupation was categorized as retired or unemployed versus all others and marital status was categorized as never married versus ever married. The multivariate logistic regression analysis forced age (as a continuous variable) and gender into the model and then used backward and forward entry for the other variables (marital status, occupation, education, urban versus rural residence, family income, religious beliefs, and prior care for psychological problems). Comparison of rates between different types of hospitals was based on relative risk (RR); the CIs for the RR uses the maximum likelihood estimate of the rate ratio [26]. Data was double-entered using and Epidata 3.1 and analyzed using SAS 8.2 and SPSS 15.0. The study was approved by the institutional review board of the China Medical University. Details of the methodology are described in a prior paper [20]. 3. Results Among the 5750 selected subjects 437 (7.6%) refused to participate and 1 only completed part of the screening process. Compared to the 5312 persons who completed the screening, the 438 noncompleters were more likely to be male (51.6% vs. 39.5%, χ2=24.45, Pb.001) and were somewhat younger (45.3±15.3 v. 48.1±17.7, t=3.55, Pb.001). The characteristics of participating subjects are shown in Table 1. The mean (S.D.) age was 48.1 (17.7) years. Over 60% were female, about 80% were currently married, over 60% had a high school education or higher and 6% came

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

195

Table 1 Characteristics of the sample and comparison of the characteristics of individuals who do and do not meet criteria of a DSM-IV anxiety disorder Characteristic

Age group 15–24 25–44 45–59 60+ Gender Male Female Marital status Never married Currently married Divorced/separated Widowed Occupationb Agricultural laborer Worker Professional Retired Unemployed Student/housewife Educational level Illiterate Elementary school High school College Residencea Urban Rural Annual income of familyb Low (b$1250) Medium ($1250–$3750) High (N$3750) Religious beliefsc No Yes Prior psychological txb No Yes

Results in all screened subjects (n=5312)

Crude frequency and adjusted % based on results of SCID exam (n=1432) Has any anxiety disorder (n=457)

Does not have any anxiety disorder (n=975)

χ2

df

P

n

%

n

Adjusted %a

n

Adjusted %a

651 1576 1576 1509

12.3 29.7 29.7 28.4

23 126 185 123

5.4 27.2 40.6 26.8

104 308 313 250

15.5 30.7 26.8 27.0

67.30

3

b.001

2100 3212

39.5 60.5

155 302

33.7 66.3

346 629

40.3 59.7

8.56

1

.003

758 4228 54 272

14.3 79.3 1.0 5.1

26 382 15 34

5.6 83.5 3.3 7.7

121 770 13 71

17.1 76.6 1.3 4.9

60.12

3

b.001

286 1162 787 2039 523 514

5.4 21.9 14.8 38.4 9.9 9.7

30 92 53 197 65 20

6.5 21.0 11.5 42.7 13.9 4.4

65 212 137 349 120 92

3.2 22.0 15.2 37.7 10.4 11.4

50.13

5

b.001

267 1409 2708 928

5.0 26.5 51.0 17.5

22 124 249 62

4.8 27.5 54.3 13.4

45 286 490 154

4.3 26.8 48.2 20.7

16.48

3

.001

4986 322

93.9 6.1

430 27

94.3 5.7

897 78

96.0 4.0

3.55

1

.059

1356 3312 643

25.5 62.3 12.1

147 248 62

32.4 54.2 13.4

319 545 110

24.1 62.3 13.7

18.06

2

b.001

5096 212

95.9 4.1

432 25

94.3 5.7

929 45

97.3 2.7

14.82

1

b.001

5266 45

99.2 0.8

442 15

96.7 3.3

955 20

99.1 0.9

22.67

1

b.001

a Adjusted rates were weighted by the proportion of SCID examinations completed in the high-risk (weight=1.13), medium-risk (3.13) and low-risk (12.00) subjects. b One case has missing data. c Four cases have missing data.

from rural areas. Only 4% reported any type of religious beliefs and less than 1% had ever received treatment for a psychological problem. Among the 1612 individuals selected for the goldstandard diagnostic assessment with SCID, 1432 (88.8%) completed the SCID, 176 (10.9%) refused and 4 (0.3%) only completed part of the examination. There were no significant differences by gender, age, educational level or economic status between the completers and noncompleters. Among the 1432 individuals who completed the SCID, 610 had a single current (one-month) diagnosis, 241 had two diagnoses and 12 had three diagnoses. Among the 253

individuals with multiple diagnoses, 216 (85%) had comorbid anxiety and mood disorders. The screening process was highly sensitive: 845 of the 863 individuals (97.9%) with any DSM-IV diagnosis were screened as at high risk for a mental illness. Among the 457 with an anxiety disorder diagnosis, the anxiety disorder was the primary diagnosis in 252 individuals and a secondary diagnosis in 205 individuals. More than one mental disorder was present In 234 (52%) of the 457 subjects with anxiety disorders; in 90% (210/234) of these individuals, the anxiety disorder was comorbid with a depressive disorder.

196

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

Table 2 Prevalence of anxiety disorders in internal medicine outpatient departments in 23 general hospitals in Shengyang, Chinaa (primary and secondary diagnoses considered)

Completed screening Completed SCID examination Any current DSM-IV diagnosis Any anxiety disorder Panic disorder Social phobia Specific phobias Obsessive compulsive disorder Posttraumatic stress disorder GAD Anxiety due to general medical condition Anxiety disorder, NOS a b c

Adjusted prevalence (%)

95% CIb

71.39 8.33 0.63 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

19.29 9.83 0.13 0.04 0.30 0.11 0.02 2.44 0.55

17.91 8.96 0.06 0.01 0.18 0.04 0.00 2.03 0.38

20.78 10.79 0.28 0.17 0.50 0.25 0.15 2.92 0.81

100.00 1.32 0.41 3.05 1.12 0.20 24.82 5.60

0.00

6.27

5.58

7.05

63.78

All screened subjects

Screened as high-risk

Screened as moderaterisk

Screened as low-risk

N

%

n

%

n

%

n

%

5312 1432 863 457 6 2 14 5 1 115 26

100.00 26.96 60.27 31.91 0.42 0.14 0.98 0.35 0.07 8.03 1.82

1107 984 845 453 6 2 14 5 1 115 26

20.84 88.89 85.67 46.04 0.61 0.20 1.42 0.51 0.10 11.69 2.64

413 132 16 4 0 0 0 0 0 0 0

7.77 31.96 12.12 3.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00

3792 316 2 0 0 0 0 0 0 0 0

289

20.18

285

28.96

4

3.03

0

% All subjects with anxiety disordersc

Only considering primary current diagnosis. 95% CI computed using the Gaussian approximation to the log likelihood of the log (rate) [23]. Based on the proportion of the overall adjusted prevalence attributed to the condition

The prevalence estimates for the various anxiety disorders are shown in Table 2. Among persons with mental disorders, 50% had an anxiety disorder either as the primary or secondary diagnosis. Among those with anxiety disorders, 64% had anxiety NOS and 25% had GAD; the other specific anxiety disorders were relatively uncommon. Considering only the primary diagnosis, the prevalences of any anxiety disorder, anxiety NOS and GAD were 4.9% (4.3–5.6%), 2.9% (2.5–3.4%) and 1.5% (1.2–1.9%), respectively.

Table 3 shows the comparison of the prevalence of anxiety disorders in the three different types of hospitals. With the exception of anxiety due to a General Medical Condition—which was more common in primary level hospitals—all other types of anxiety disorders were more prevalent in patients seen at tertiary hospitals than at primary or secondary hospitals. The higher prevalence in tertiary versus secondary hospitals was significantly different for specific phobias, GAD and any anxiety disorder.

Table 3 Comparison of prevalence of anxiety disorders in internal medicine outpatient departments of three different types of general hospitals in Shenyang, China Diagnosis

Ten primary hospitals

Seven secondary hospitals

Six tertiary hospitals

Primary vs. secondary hospitals

Primary vs. tertiary hospitals

Secondary vs. tertiary hospitals

% (95% CIa)

% (95% CIa)

% (95% CIa)

RR (95% CIb)

RR (95% CIb)

RR (95% CIb)

Any DSM-IV anxiety disorder on SCID Panic disorder

9.88 (7.86–12.40) 0.00

11.61 (10.29–13.15)

1.30 (0.99–1.72) –

0.85 (0.66–1.10) –

Social phobia

0.00

7.57 (6.38–8.99) 0.06 (0.01–0.41) 0.00

–

–

0.65c (0.53–0.81) 0.25 (0.03–2.15) –

Specific phobias

0.24 (0.06–0.96) 0.00

0.06 (0.00–0.41) 0.00

4.11 (0.37–45.33) –

0.47 (0.10–2.12) –

0.11c (0.01–0.89) –

0.00

0.00

–

–

–

2.51 (1.64–3.86) 1.08 (0.56–2.07) 6.04 (4.47–8.17)

1.28 (0.84–1.95) 0.52 (0.27–1.01) 5.65 (4.63–6.90)

1.98c (1.08–3.57) 2.06 (0.82–5.18) 1.07 (0.76–1.52)

0.75 (0.46–1.23) 2.91c (1.12–7.54) 0.88 (0.64–1.23)

0.38c (0.24–0.62) 1.42 (0.55–3.67) 0.83 (0.64–1.07)

Obsessive compulsive disorder Posttraumatic stress disorder GAD Anxiety due to general medical condition Anxiety disorder NOS a b c

0.23 (0.10–0.56) 0.09 (0.02–0.37) 0.51 (0.28–0.92) 0.23 (0.10–0.56) 0.05 (0.00–0.33) 3.33 (2.65–4.20) 0.37 (0.12–0.74) 6.84 (5.80–8.06)

95% CI computed using the Gaussian approximation to the log likelihood of the log (rate) [24]. CI for RR uses the maximum likelihood estimate of the rate ratio [24]. Statistically significant difference.

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

197

were recognized by treating clinicians. None of the cases were both diagnosed and treated for anxiety, but in three of the six cases in which the clinician diagnosed anxiety, the patient also had a comorbid depressive disorder (both on the SCID and according to the internist's record) and the patient was treated with an antidepressant. Considering an anxiety diagnosis or anxiolytic treatment as “recognition,” the overall recognition rate for anxiety disorders is 4.0% (16/ 402, 95% CI=2.3–6.3%). Misdiagnosis of anxiety as depression was rare; there was only one case in which a single clinical diagnosis of depression was assigned to a patient with a SCID-based anxiety disorder, but the SCID exam for this patient also identified a major depressive disorder and alcohol abuse in addition to anxiety NOS.

The last four columns of Table 1 compare characteristics of the 457 identified subjects with anxiety disorders to those of the 975 subjects who did not have anxiety disorders. After weighting and dichotomization of the multilevel categorical variables, univariate analysis found that outpatients with anxiety disorders were older than those without anxiety disorders [mean (S.D.) age=50.1 (14.7) vs. 46.5 (18.3) years; t=4.29, Pb.001] and were significantly more likely to be female; currently married, divorced or widowed (94.4% vs. 82.9%; χ2=46.6, Pb.001); retired or unemployed (56.6% vs. 48.1%; χ2=14.5, Pb.001); to have less than a college education (86.6% vs. 79.3%; χ2=15.5, Pb.001); to have a low family income (32.4% vs. 24.1%, χ2=17.4, Pb.001); to have religious beliefs and to have previously sought help for psychological problems. All of these factors were entered into a multivariate logistic regression analysis. After forcing age (as a continuous variable) and gender into the model, backward and forward entry of the remaining variables identified five independent predictors of having an anxiety disorder (in order of importance): every being married (OR=3.5, 95% CI=2.3–5.4), prior treatment for psychological problems (3.3, 1.8–6.0), having religious beliefs (1.9, 1.3–2.9), low family income (1.5, 1.2–1.9) and never having attended college (1.3, 1.02–.8) (neither age nor gender remained significant in the final model). Recognition rates for the different anxiety disorders are presented in Table 4. Only six (1.5%) of the 402 patients with SCID-based anxiety disorders whose charts were reviewed were given an anxiety disorder diagnosis by the treating internist, and a further 10 (2.5%) were treated with anxiolytic medications without being assigned any diagnosis. Anxiety disorder NOS and GAD (the most prevalent conditions) were the only types of anxiety disorders that

4. Discussion This study has several strengths. It is the largest study about the prevalence and recognition rates of anxiety in internal medicine clinics of general hospitals yet available in China. The sample is representative of all internal medicine outpatients at general hospitals in a large urban municipality in northern China, the refusal rate was quite low, the gold standard diagnosis was based on the administration of SCID by psychiatrists who had excellent inter-rater reliability, the assessment of recognition rates was based on clinical records rather than on directly asking internists about the presence of anxiety (which could inflate the recognition rates), and rigorous quality control measures were enforced throughout the study. The health care delivery system in Shenyang is similar to that in other urban areas of China, so we believe that the results are

Table 4 Recognition of anxiety disorders by treating clinicians in internal medicine outpatient departments of 23 general hospitals in Shenyang, China Diagnosis

Any DSM-IV anxiety disorder on SCIDa Panic disorder Specific phobias Obsessive compulsive disorder Posttraumatic stress disorder GAD Anxiety due to general medical condition Anxiety disorder NOS a

Subjects with gold-standard diagnosis (SCID)

Clinician recorded anxiety diagnosis in chart but gave no treatment

Clinician treated with anxiolytics but recorded no diagnosis in chart

Clinician diagnosed anxiety and treated with anxiolyticsa

All identified subjects (diagnosis in chart or treated)

n

% (95% CI)

n

% (95% CI)

n

% (95% CI)

n

% (95% CI)

402

6

10

0.0

16

0 0 0

0 0 0

2.5% (1.2–4.5%) 0.0 0.0 0.0

0

1 3 4

1.5% (0.5–3.1%) 0.0 0.0 0.0

0 0 0

0.0 0.0 0.0

0 0 0

4.0% (2.3–6.3%) 0.0 0.0 0.0

1

0

0.0

0

0.0

0

0.0

0

0.0

106

2

2

0.0

4

0

1.9% 0.2–6.7%) 0.0

0

22

1.9% (0.2–6.7%) 0.0

0

0.0

0

3.8% (1.0–9.4%) 0.0

265

4

0

0.0

12

1.5% (0.4–3.8%)

0 8

3.0% (1.3–5.9%)

4.5% (2.4–7.8%)

Two subjects with anxiety disorder NOS and one subject with GAD who had concurrent major depressive disorder were treated with anti-depressants.

198

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

relevant to other parts of urban China and may also be relevant to urban parts of other low- and middle-income countries that have a similar “collective-care” model for providing primary care services—in which patients are often seen by different clinicians at each visit and may not be interviewed or treated in a private examination room. 4.1. Prevalence and characteristics of anxiety in primary care in urban China The 9.8% current prevalence of any anxiety disorder in internal medicine outpatients identified in this study is lower than that reported in most prior studies in primary care settings, which range from 11% to 19% [27–30]. One reason for the lower prevalence is that most other studies report 1year prevalence, while our study reports 1-month prevalence. However, our relatively low prevalence is consistent with the results of the 1991 international “Mental Illness in General Health Care” study [31] which reported a current prevalence of GAD in Shanghai (the only Chinese site included in the study) that was the second lowest of all 15 participating countries (1.9% compared to the range of 0.9–18.7% at all sites). The lower reported prevalence could be due to (1) an overall lower prevalence of anxiety disorders in mainland Chinese residents, (2) a less frequent use of general health care services by persons with anxiety disorders or (3) a greater reluctance of Chinese attendees to general health care services to report anxiety symptoms. The relatively high prevalence of anxiety disorder NOS (6.3%) and the high proportion of all subjects with anxiety diagnoses who had this diagnosis (64%) was also found in two other studies that considered this subsyndromal diagnosis. These studies, which employed the Primary Care Evaluation of Mental Disorders [32] to assess DSMIV diagnoses, reported that 11.4% [18] and 8.5% [27] of primary care attendees meet anxiety disorder NOS criteria and that 77% [18] and 45% [27] of all subjects with anxiety disorders had the NOS diagnosis. There are two primary explanations for this finding. First, patients with sub-threshold anxiety symptoms are presumed to be common among persons with physical ailments. Second, patients who do not experience or describe their anxiety symptoms in ways that meet standardized definitions of the threshold anxiety disorders (phobia, panic, GAD, PTSD, etc.) will quite often be relegated to the NOS category. Most studies about anxiety in primary care do not consider this diagnosis, presumably because of doubts about the value of identifying patients with subsyndromal anxiety symptoms or concerns about medicalizing normal distress. The fundamental question about whether or not individuals with the anxiety disorder NOS diagnosis are sufficiently distressed or disabled by their symptoms to merit evaluation, monitoring and, perhaps, treatment by the primary care clinician will not be resolved until there is more research on the long-term outcomes of primary care attendees with this diagnosis.

The high rate of comorbidity of anxiety disorders with other mental disorders (52%) found in our study was consistent with the 21–70% comorbidity reported in prior studies [18,27–29] in primary care settings. The high proportion of these dual diagnosis subjects who had comorbid anxiety and depressive disorders (90%) was also found in other studies [12]. This result confirms the existence of the “mixed anxiety-depressive” diagnostic entity described in the appendix of the DSM-IV. Few studies have looked at risk factors for the general category of anxiety disorders in primary care so it is not possible to directly compare our risk-factor results with other research. However, studies on anxiety in the community [33] typically report that anxiety disorders are more common in women, in younger adults, in the unmarried or divorced and in persons of low educational or economic status. Our univariate results confirmed the association of anxiety disorder with female gender and low economic or educational status, but contrary to previous results, we found that among primary care attendees in China young adults and those who were unmarried were less likely to suffer from an anxiety disorder and those who report religious beliefs are more likely to have an anxiety disorder. After adjusting for prior mental health treatment in the multivariate analysis, gender and age were not found to be independently related to anxiety disorders but the protective effect of being unmarried or having a college education and the risk associated with having a low family income or having religious beliefs persisted. This finding highlights the importance of considering the ‘cultural valence’ of different risk factors in crossnational studies. For example, unlike in the west where religious beliefs are common and typically transmitted through families, very few urban Chinese have religious beliefs and most who do have such beliefs are recent converts who have sought out religion because of a desire to find more meaning in their lives. Similarly, the personal meaning of marriage varies greatly by culture so the emotionally protective effect of marriage seen in western nations may not be as relevant in mainland China. 4.2. Recognition and treatment of anxiety disorders in primary care in urban China The very low rate of recognition of anxiety disorders by internists (4.0%) is identical to our previously reported 4.0% rate of recognition of depressive disorders [20]. Recognition rates for any anxiety disorder or for GAD in primary care in western studies range from 14% to 73% [12,15,30,34], but some of these studies use other diagnostic criteria and most of them asked treating clinicians to record the presence or absence of anxiety, so these recognition rates are not directly comparable with the current study. However, our low recognition rate is also quite similar to the 5.4% recognition rate for any mental disorder reported from Shanghai in the 1991 international “Mental Illness in General Health Care” study [31]. In that study—which used a similar methodology

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

199

across sites—the Shanghai recognition rate was the second lowest among the 15 participating sites (the recognition rates varied from 4.8% to 58.5% across the sites). Thus, it seems reasonable to conclude that recognition rates for anxiety (and other mental disorders) are substantially lower in urban Chinese primary care settings than in similar settings in most other countries. There are several reasons for the low recognition of anxiety disorders: (1) patients' reluctance to report psychological symptoms, (2) clinicians are not trained to assess or treat mental disorders, (3) clinicians' reluctance to give a stigmatizing diagnosis, (4) the brief clinical encounter, (5) lack of privacy in the clinical encounter and (6) lack of a designated primary care provider (i.e., personal physician) so patients see different clinicians at each clinic visit. Resolution of these problems will be a long-term process that must simultaneously focus on two complex issues: (1) providing education about mental disorders and changing attitudes about psychological problems both among community members and among health professionals, and (2) changing the method of delivering primary care services in ways that will increase the recognition of mental disorders, for example, by screening all outpatients and assigning high-risk individuals to designated providers who have specialized training in the recognition and management of mental disorders. Such major changes require substantial resources and sustained effort so they will remain unachieved objectives until the political will for changing the care delivery system has been mobilized and until the incentive system for clinicians is changed to ensure that they are motivated to identify and treat patients with psychological problems.

illness (26.8% vs. 32.1%, χ =16.53, Pb.001), and the final estimated prevalence of any anxiety disorder in those screened in private rooms versus those screened in the open clinic were quite similar (9.60% versus 10.41%, RR=1.08, 95% CI=0.90–1.31). We did not record outpatient clinicians' diagnoses and treatment among patients who did not meet the gold standard diagnosis of a mental disorder, so it was not possible to determine whether or not treating clinicians inappropriately diagnosed anxiety or provided anxiolytics to persons without a diagnosis (i.e., the false-positive rate). But given the extremely low rates of care-seeking for psychological problems (1%) and the low rates of diagnosis and treatment in persons who did meet gold standard diagnostic criteria, it is unlikely that there were substantial numbers of patients without a gold standard diagnosis who were given a diagnosis of anxiety or treated with anxiolytics by the internists. Our method of estimating recognition rates could underestimate the actual recognition rate if clinicians (despite the very short contact time) recognize anxiety without recording the diagnosis or providing treatment, but in the absence of any intervention, such unrecorded recognition is unlikely to have an effect on the course of the illness for the patient so it is, for all practical purposes, equivalent to nonrecognition. We believe that the problem of underreporting recognition rates by depending on clinical records is much less serious than the problem of inflating recognition rates by requiring the treating internists to evaluate each patient in the study for anxiety.

4.3. Limitations

This project was part of the “Small Grants Program to Improve the Quality and Implementation of Mental Health Research in China” which was supported by the China Medical Board of New York (Grant Number 02–777) and coordinated by Professor Michael Phillips of the Beijing Hui Long Guan Hospital and Professor Xue Zhang of Peking Union Medical College. We would like to thank the 23 participating hospitals in Shenyang for their active support of the project and Xianyun Li, Zhiqing Wang, Yali Zhang, Yajuan Niu from the Beijing Suicide Research and Prevention Center at the Beijing Hui Long Guan Hospital for their assistance in coordinating the project.

Several issues need to be considered when interpreting our results. Only a small proportion of identified subjects (7.6%) refused to participate, but those who did refuse were younger and more likely to be male. However, there is no compelling reason to suspect that patients who refused have higher rates of anxiety, so it is unlikely that the low refusal rate seriously compromised the representativeness of the sample or validity of the overall results. The sample size was sufficient to provide reasonably precise estimates of the prevalence of the different anxiety disorders, but the very small numbers of subjects with anxiety disorders recognized by internists resulted in large CIs for the recognition rates. Outpatient departments in urban Chinese general hospitals are often crowded and noisy, so it is difficult to provide a suitable environment to conduct a psychological interview. In 35% of the GHQ screening interviews, we were unable to secure a separate clinic room so respondents may have been more reluctant to subscribe to anxiety symptoms. However, contrary to our expectation, comparison of the GHQ results found that subjects interviewed in private rooms were less likely to be classified as moderate to high risk of a mental

2

Acknowledgments

References [1] Murray CJL, Lopez AD. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge (MA): Harvard University Press; 1996. [2] Murray CJL, Lopez AD. Global health statistics: a compendium of incidence, prevalence, and mortality estimates for over 200 conditions. Cambridge (MA): Harvard University Press; 1996. [3] WHO. The global burden of disease: 2004 update. Geneva: World Health Organization; 2008.

200

X. Qin et al. / General Hospital Psychiatry 32 (2010) 192–200

[4] Kessler RC. The global burden of anxiety and mood disorders: putting ESEMeD findings into perspective. J Clin Psychiatry 2007;68 (suppl 2):10–9. [5] Tyrer P. Are general practitioners really unable to diagnose depression? [comment]. Lancet 2009;374:589–90. [6] Bakker IM, Terluin B, van Marwijk HW, van Mechelen W, Stalman WA. Test-retest reliability of the PRIME-MD: limitations in diagnosing mental disorders in primary care. Eur J Public Health 2009;19:303–7. [7] Todd C, Patel V, Simunyu E, Gwanzura F, Acuda W, Winston M, et al. The onset of common mental disorders in primary care attenders in Harare, Zimbabwe. Psychol Med 1999;29:97–104. [8] Patel VH, Kirkwood BR, Pednekar S, Araya R, King M, Chisholm D, et al. Improving the outcomes of primary care attenders with common mental disorders in developing countries: a cluster randomized controlled trial of a collaborative stepped care intervention in Goa, India. Trials 2008;9:4. [9] Liu SI, Huang HC, Yeh ZT, Hwang LC, Tjung JJ, Huang CR, et al. Controlled trial of problem-solving therapy and consultation-liaison for common mental disorders in general medical settings in Taiwan. Gen Hosp Psychiatry 2007;29:402–8. [10] Lancet Global Mental Health Group. Scale up services for mental disorders: a call for action. Lancet 2007;370:1241–52. [11] WHO. Mental Health Gap Action Program (mhGAP): scaling up care for mental, neurological, and substance use disorders. Geneva: World Health Organization; 2008. [12] Wittchen HU, Kessler RC, Beesdo K, Krause P, Höfler M, Hoyer J. Generalized anxiety and depression in primary care: prevalence, recognition, and management. J Clin Psychiatry 2002;63(Suppl 8): 24–34. [13] Olssøn I, Mykletun A, Dahl AA. Recognition and treatment recommendations for generalized anxiety disorder and major depressive episode: a cross-sectional study among general practitioners in Norway. J Clin Psychiatry 2006;8:340–7. [14] Brenes GA, Knudson M, McCall WV, Williamson JD, Miller ME, Stanley MA. Age and racial differences in the presentation and treatment of generalized anxiety disorder in primary care. J Anxiety Disor 2008;22:1128–36. [15] Roy-Byrne PP, Wagner A. Primary care perspectives on generalized anxiety disorder. J Clin Psychiatry 2004;65(Suppl 13):20–6. [16] Liebschutz J, Saitz R, Brower V, Keane TM, Lloyd-Travaglini C, Averbuch T, et al. PTSD in urban primary care: high prevalence and low physician recognition. J Gen Intern Med 2007;22:888–9. [17] Yang YK, Yeh TL, Chen CC, Lee CK, Lee IH, Lee LC, et al. Psychiatric morbidity and posttraumatic symptoms among earthquake victims in primary care clinics. Gen Hosp Psychiatry 2003;25:253–61. [18] Jackson JL, Passamonti M, Kroenke K. Outcome and impact of mental disorders in primary care at 5 years. Psychosom Med 2007;69:270–6. [19] Phillips MR, Zhang JX, Shi QC, Song ZQ, Ding ZJ, Pang ST, et al. Prevalence, associated disability and treatment of mental disorders in

[20]

[21]

[22]

[23]

[24]

[25]

[26] [27]

[28] [29]

[30]

[31]

[32]

[33]

[34]

four provinces in China, 2001-2005: an epidemiological survey. Lancet 2009;373:2041–53. Qin XX, Wang W, Jin Q, Ai L, Li YL, Dong GH, et al. Prevalence and rates of recognition of depressive disorders in internal medicine outpatient departments of 23 general hospitals in Shenyang, China. J Affect Dis 2008;110:46–54. Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, Gureje O, et al. The validity of two versions of the GHQ study of mental illness in general health care. Psychol Med 1997;27:191–7. First MB, Spitzer RL, Gibbon M, Williams JB. Structured Clinical Interview for DSM-IV-TR Axis I Disorders. New York: Biometrics Research Department, New York State Psychiatric Institute; 2002. Yang TZ, Huang L, Wu ZY. Study on the appropriateness of the Chinese version of the General Health Questionnaire as a screening instrument for psychological disorders in mainland China. Chin J Epid 2003;24:769–73 (in Chinese). Zhou RY, Zhang YH, Peng B, Lie XH, Zhu CM. Comparison of three diagnostic criteria for the diagnosis of schizophrenia and mood disorders. Chin J Psychiatry 1997;30:45–9 (in Chinese). Shi QC, Zhang JM, Xu FZ, Phillips MR, Xu Y, Fu YL, et al. Epidemiological survey of mental illnesses in Zhejiang Province, China. Chin J Prev Med 2005;229-236:39 (in Chinese). Clayton D, Hills M. Statistical models in epidemiology. Oxford: Oxford Science Publications, 1993.73–88,133–140. Ansseau M, Dierick M, Buntinkx F, Cnockaert P, De Smedt J, Van Den Haute M, et al. High prevalence of mental disorders in primary care. J Affect Disord 2004;78:49–55. Nisenson LG, Pepper CM. The nature and prevalence of anxiety disorders in primary care. Gen Hosp Psychiatry 1998;20:21–8. Cwikel J, Zilber N, Feinson M, Lerner Y. Prevalence and risk factors of threshold and sub-threshold psychiatric disorders in primary care. Soc Psychiatry Psychiatr Epidemiol 2008;43:184–91. Furedi J, Rozsa S, Zambori J, Szadoczky E. The role of symptoms in the recognition of mental health disorders in primary care. Psychosomatics 2003;44:402–6. Ustun TB, Sartorius N (Eds). Mental Illness in General Health Care: an international study. Chichester, England: John Wiley & Son. 1995: 285–300, 345–369. Spitzer RL, Williams JB, Kroenke K, Linzer M, deGruy FVIII, Hahn SR, et al. Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study. JAMA 1994;272:1749–56. Alonso J, Angermeyer MC, Bernert S, Bruffaerts R, Brugha TS, Bryson H, et al. Prevalence of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatr Scand 2004;420(Suppl):21–7. Munk-Jørgensen P, Allgulander C, Dahl AA, Foldager L, Holm M, Rasmussen I, et al. Prevalence of generalized anxiety disorder in general practice in Denmark, Finland, Norway, and Sweden. Psychiatr Serv 2006;57:1738–44.